TreeBASE Search-Matrices for study 1733

@ARTICLE{TreeBASE2Ref17862, author = {Erin Tripp}, title = {Evolutionary relationships within the species-rich genus Ruellia (Acanthaceae)}, year = {2007}, keywords = {}, doi = {}, url = {}, pmid = {}, journal = {Systematic Botany}, volume = {}, number = {}, pages = {}, abstract = {Phylogenetic relationships among species of the large genus Ruellia (Acanthaceae) have never been studied. Ruellia, with approximately 300 species, is geographically widespread and morphologically diverse. Molecular data for 196 specimens from the nuclear ribosomal ITS region and the chloroplast trnG-trnR region were used to test monophyly of the genus against closely related genera in Ruellieae, to reconstruct phylogenetic relationships among species of Ruellia on a global scale, to re-evaluate previous morphology-based classifications, and to examine the utility of morphological characters, especially corolla morphology, for future sectional delimitation. Bayesian and parsimony analyses indicate four genera are evolutionarily allied to Ruellia. Acanthopale is sister to Ruellia s.l. with strong support. Blechum, Eusiphon, and Polylychnis are nested within Ruellia s.l., and species in Eusiphon and Polylychnis are here formally transferred to Ruellia resulting in the new combinations Ruellia geayi and Ruellia fulgens. Ruellia s.l., including Blechum, Eusiphon, and Polylychnis, is monophyletic but only weakly supported by parsimony. Within Ruellia, Old World taxa form a basal grade and New World taxa are monophyletic and nested within the Old World grade. Alternative hypotheses involving non-monophyly of New World Ruellia were significantly less parsimonious and less likely. Within New World Ruellia, many clades are informally recognized, several of which reflect previous taxonomic groupings to some extent. An alternative hypothesis constraining all putatively hummingbird-pollinated taxa to monophyly was strongly rejected suggesting that corolla morphology has undergone convergent evolution and is therefore likely an inappropriate character for sectional delimitation, contrary to previous use.} }